Commit Graph

1453 Commits

Author SHA1 Message Date
Chris Lattner
3749c9cae4 Remove the 'printname' argument to WriteAsOperand. It is always true, and
passing false would make the asmprinter fail anyway.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32264 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-06 06:16:21 +00:00
Chris Lattner
e13ab2a9a1 add an instcombine xform. This speeds up 462.libquantum from 9.78s to
7.48s.  This regression is due to unforseen consequences of the cast patch.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32209 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-05 01:26:29 +00:00
Devang Patel
67a821d9a3 SCCP does not handle Packed Type properly. Disable Packed Type handling
for now.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32208 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-04 23:54:59 +00:00
Reid Spencer
3b3844288b Update call to CastInst::getCastOpcode for its new signature.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32166 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-04 02:48:01 +00:00
Jeff Cohen
97af751deb Unbreak VC++ build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32113 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-02 02:22:01 +00:00
Chris Lattner
9919e3df17 disable transformations that are invalid for fp vectors. This fixes
Transforms/InstCombine/2006-12-01-BadFPVectorXform.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32112 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-02 00:13:08 +00:00
Reid Spencer
44c030a7ee Remove 4 FIXMEs to hack around cast-to-bool problems which no longer exist.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32051 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-30 23:13:36 +00:00
Chris Lattner
46b96055cf implement cast.ll:test35. With this, we recognize:
unsigned short swp(unsigned short a) {
       return ((a & 0xff00) >> 8 | (a & 0x00ff) << 8);
}

as an idiom for bswap.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32011 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-29 07:18:39 +00:00
Chris Lattner
6aa5eb19d5 Teach instcombine to turn trunc(srl x, c) -> srl (trunc(x), c) when safe.
This implements InstCombine/cast.ll:test34.  It fires hundreds of times on
176.gcc.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32009 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-29 07:04:07 +00:00
Chris Lattner
458cf462ef Implement Regression/Transforms/InstCombine/bswap-fold.ll,
folding   seteq (bswap(x)), c -> seteq(x,bswap(c))


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32006 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-29 05:02:16 +00:00
Reid Spencer
de33124aa3 Join a split line.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31996 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-29 01:11:01 +00:00
Reid Spencer
d4d9ab80b7 Undo the last patch until 253.perlbmk passes with these changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31977 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-28 20:23:51 +00:00
Reid Spencer
5d4dbd1f60 Remove 4 FIXME's from the CAST patch now that the back end is correctly
producing code for "trunc to bool". This passes all tests on Linux.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31963 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-28 07:23:01 +00:00
Chris Lattner
f2f1643037 Fix PR1014 and InstCombine/2006-11-27-XorBug.ll.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31941 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-27 19:55:07 +00:00
Reid Spencer
3da59db637 For PR950:
The long awaited CAST patch. This introduces 12 new instructions into LLVM
to replace the cast instruction. Corresponding changes throughout LLVM are
provided. This passes llvm-test, llvm/test, and SPEC CPUINT2000 with the
exception of 175.vpr which fails only on a slight floating point output
difference.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31931 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-27 01:05:10 +00:00
Bill Wendling
b742703137 Removed #include <iostream> and replaced with llvm_* streams.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31923 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-26 09:46:52 +00:00
Nick Lewycky
565706b93e Update to new predicate simplifier VRP design. Fixes PR966 and PR967.
Remove predicate simplifier from default gcc3 pipeline. New design is too
slow to enable by default.
Add new testcases for problems encountered in development.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31895 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-22 23:49:16 +00:00
Chris Lattner
13c654a42a This xform is handled by FoldOpIntoPhi in visitCastInst in a more elegant way.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31889 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-21 17:05:13 +00:00
Chris Lattner
020f714a93 If an indvar with a variable stride is used by the exit condition, go ahead
and handle it like constant stride vars.  This fixes some bad codegen in
variable stride cases.  For example, it compiles this:

void foo(int k, int i) {
  for (k=i+i; k <= 8192; k+=i)
    flags2[k] = 0;
}

to:

LBB1_1: #bb.preheader
        movl %eax, %ecx
        addl %ecx, %ecx
        movl L_flags2$non_lazy_ptr, %edx
LBB1_2: #bb
        movb $0, (%edx,%ecx)
        addl %eax, %ecx
        cmpl $8192, %ecx
        jle LBB1_2      #bb
LBB1_5: #return
        ret

or (if the array is local and we are in dynamic-nonpic or static mode):

LBB3_2: #bb
        movb $0, _flags2(%ecx)
        addl %eax, %ecx
        cmpl $8192, %ecx
        jle LBB3_2      #bb

and:

        lis r2, ha16(L_flags2$non_lazy_ptr)
        lwz r2, lo16(L_flags2$non_lazy_ptr)(r2)
        slwi r3, r4, 1
LBB1_2: ;bb
        li r5, 0
        add r6, r4, r3
        stbx r5, r2, r3
        cmpwi cr0, r6, 8192
        bgt cr0, LBB1_5 ;return

instead of:

        leal (%eax,%eax,2), %ecx
        movl %eax, %edx
        addl %edx, %edx
        addl L_flags2$non_lazy_ptr, %edx
        xorl %esi, %esi
LBB1_2: #bb
        movb $0, (%edx,%esi)
        movl %eax, %edi
        addl %esi, %edi
        addl %ecx, %esi
        cmpl $8192, %esi
        jg LBB1_5       #return

and:

        lis r2, ha16(L_flags2$non_lazy_ptr)
        lwz r2, lo16(L_flags2$non_lazy_ptr)(r2)
        mulli r3, r4, 3
        slwi r5, r4, 1
        li r6, 0
        add r2, r2, r5
LBB1_2: ;bb
        li r5, 0
        add r7, r3, r6
        stbx r5, r2, r6
        add r6, r4, r6
        cmpwi cr0, r7, 8192
        ble cr0, LBB1_2 ;bb

This speeds up Benchmarks/Shootout/sieve from 8.533s to 6.464s and
implements LoopStrengthReduce/var_stride_used_by_compare.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31809 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-17 06:17:33 +00:00
Chris Lattner
d23b5baf8a Fix a gcc 4.2 warning.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31751 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-15 04:53:24 +00:00
Chris Lattner
e511b74f4a implement InstCombine/shift-simplify.ll by transforming:
(X >> Z) op (Y >> Z)  -> (X op Y) >> Z

for all shifts and all ops={and/or/xor}.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31729 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-14 07:46:50 +00:00
Chris Lattner
9c2328e5a9 implement InstCombine/and-compare.ll:test1. This compiles:
typedef struct { unsigned prefix : 4; unsigned code : 4; unsigned unsigned_p : 4; } tree_common;
int foo(tree_common *a, tree_common *b) { return a->code == b->code; }

into:

_foo:
        movl 4(%esp), %eax
        movl 8(%esp), %ecx
        movl (%eax), %eax
        xorl (%ecx), %eax
        # TRUNCATE movb %al, %al
        shrb $4, %al
        testb %al, %al
        sete %al
        movzbl %al, %eax
        ret

instead of:

_foo:
        movl 8(%esp), %eax
        movb (%eax), %al
        shrb $4, %al
        movl 4(%esp), %ecx
        movb (%ecx), %cl
        shrb $4, %cl
        cmpb %al, %cl
        sete %al
        movzbl %al, %eax
        ret

saving one cycle by eliminating a shift.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31727 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-14 06:06:06 +00:00
Chris Lattner
de2b6605ba Fix InstCombine/2006-11-10-ashr-miscompile.ll a miscompilation introduced
by the shr -> [al]shr patch.  This was reduced from 176.gcc.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31653 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-10 23:38:52 +00:00
Chris Lattner
b4a2f059ad Teach ShrinkDemandedConstant how to handle X+C. This implements:
add.ll:test33, add.ll:test34, shift-sra.ll:test2


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31586 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-09 05:12:27 +00:00
Chris Lattner
53738a4dbf reenable factoring of GEP expressions, being more precise about the
case that it bad to do.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31563 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-08 19:42:28 +00:00
Chris Lattner
f6fd94d3fa make this code more efficient by not creating a phi node we are just going to
delete in the first place.  This also makes it simpler.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31562 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-08 19:29:23 +00:00
Chris Lattner
f0cd7085c1 disable this factoring optzn for GEPs for now, this severely pessimizes some
loops.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31560 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-08 18:49:31 +00:00
Reid Spencer
3822ff5c71 For PR950:
This patch converts the old SHR instruction into two instructions,
AShr (Arithmetic) and LShr (Logical). The Shr instructions now are not
dependent on the sign of their operands.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31542 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-08 06:47:33 +00:00
Chris Lattner
d92515034f scalarrepl should not split the two elements of the vsiidx array:
int func(vFloat v0, vFloat v1) {
        int ii;
        vSInt32 vsiidx[2];
        vsiidx[0] = _mm_cvttps_epi32(v0);
        vsiidx[1] = _mm_cvttps_epi32(v1);
        ii = ((int *) vsiidx)[4];
        return ii;
}

This fixes Transforms/ScalarRepl/2006-11-07-InvalidArrayPromote.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31524 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-07 22:42:47 +00:00
Jeff Cohen
d41b30def3 Unbreak VC++ build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31464 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-05 19:31:28 +00:00
Nick Lewycky
8b2989ae91 Remove commented line from earlier debugging.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31460 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-05 14:19:40 +00:00
Andrew Lenharth
8ed4c47d70 The wrong parameter was being tested to deturmine i32 vs i64
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31431 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-03 22:45:50 +00:00
Chris Lattner
ffa0a9b7ea remove dead code
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31398 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-03 01:34:58 +00:00
Reid Spencer
3ed469ccd7 For PR786:
Turn on -Wunused and -Wno-unused-parameter. Clean up most of the resulting
fall out by removing unused variables. Remaining warnings have to do with
unused functions (I didn't want to delete code without review) and unused
variables in generated code. Maintainers should clean up the remaining
issues when they see them. All changes pass DejaGnu tests and Olden.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31380 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-02 20:25:50 +00:00
Reid Spencer
0a783f783c For PR950:
Replace the REM instruction with UREM, SREM and FREM.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31369 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-02 01:53:59 +00:00
Devang Patel
e0805a2d36 There can be more than one PHINode at the start of the block.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31362 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-01 23:04:45 +00:00
Devang Patel
2c0565f8e0 Handle PHINode with only one incoming value.
This fixes http://llvm.org/bugs/show_bug.cgi?id=979


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31358 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-01 22:26:43 +00:00
Chris Lattner
9c08050ac7 Factor gep instructions through phi nodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31346 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-01 07:43:41 +00:00
Chris Lattner
76c7314650 Turn a phi of many loads into a phi of the address and a single load of the
result.  This can significantly shrink code and exposes identities more
aggressively.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31344 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-01 07:13:54 +00:00
Chris Lattner
a90a24c623 Fix a bug in the previous patch
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31342 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-01 04:55:47 +00:00
Chris Lattner
7da52b295b Fold things like "phi [add (a,b), add(c,d)]" into two phi's and one add.
This triggers thousands of times on multisource.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31341 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-01 04:51:18 +00:00
Chris Lattner
0997fadc37 break edges more intelligently
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31257 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-28 06:45:33 +00:00
Chris Lattner
f6de8ad1a2 SplitCriticalEdge checks to see if an edge is critical, don't check twice
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31255 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-28 06:38:14 +00:00
Chris Lattner
1b9c8e73b5 prepare for a change I'm about to make
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31248 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-28 00:59:20 +00:00
Reid Spencer
811b0cbd6e Simplify code a bit by changing instances of:
InsertNewInstBefore(new CastInst(Val, ValTy, Val->GetName()), I)
into:
   InsertCastBefore(Val, ValTy, I)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31204 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-26 19:19:06 +00:00
Reid Spencer
1628cec4d7 For PR950:
Make necessary changes to support DIV -> [SUF]Div. This changes llvm to
have three division instructions: signed, unsigned, floating point. The
bytecode and assembler are bacwards compatible, however.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31195 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-26 06:15:43 +00:00
Nick Lewycky
7043d00750 Fix 2006-10-25-AddSetCC. A relational operator (like setlt) can never
produce an EQ property.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31193 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-26 02:35:18 +00:00
Nick Lewycky
e63bf95a32 Resurrect r1.25.
Fix and comment the "or", "and" and "xor" transformations.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31189 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-25 23:48:24 +00:00
Chris Lattner
b67e8f2263 hide symbols properly
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31184 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-25 21:14:31 +00:00
Chris Lattner
ae5d51c9c9 Fix Transforms/ScalarRepl/2006-10-23-PointerUnionCrash.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31151 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-24 06:26:32 +00:00
Chris Lattner
ef2aa193d6 Revert back to r1.21, which was the last revision of predsimplify that
passes llvm-gcc bootstrap.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31146 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-24 00:36:21 +00:00
Chris Lattner
ddaaa37487 Handle fallout from the recent branch-on-undef changes. This fixes
Prolangs-C/agrep and SCCP/2006-10-23-IPSCCP-Crash.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31132 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-23 18:57:02 +00:00
Nick Lewycky
c17229d591 Remove the Backwards operation. Resolving now works at the time when a
property is added by running through the list of uses of the value and
adding resolved properties to the property set.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31126 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-23 01:56:02 +00:00
Nick Lewycky
7e189d095e Fix similar missing optimization opportunity in XOR.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31123 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-22 22:22:58 +00:00
Nick Lewycky
92a8b71dc1 Whoops! Add missing NULL check.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31121 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-22 21:38:24 +00:00
Nick Lewycky
5062250f36 Handle "if ((x|y) != 0)" for ints like we do for bools. Fixes missed
optimization opportunity pointed out by Chris Lattner.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31118 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-22 21:36:41 +00:00
Nick Lewycky
802fe27d3e AllocaInst can't return a null pointer. Fixes missed optimization
opportunity pointed out by Andrew Lewycky.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31115 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-22 19:53:27 +00:00
Chris Lattner
32b1e87f11 Add a workaround for PR962, disabling the more aggressive form of this
transformation.  This speeds up a C++ app 2.25x.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31113 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-22 18:42:26 +00:00
Chris Lattner
d2d86703f0 3 Changes:
1. Better document what is going on here.
2. Only hack on one branch per iteration, making the results less conservative.
3. Handle the problematic case by marking edges executable instead of by
   playing with value lattice states.  This is far less pessimistic, and fixes
   SCCP/ipsccp-gvar.ll.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31106 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-22 05:59:17 +00:00
Chris Lattner
6734bffe24 Fix an ugly problem in SCCP. This fixes Benchmarks/Misc-C++/mandel-text.cpp
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31073 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-20 20:19:08 +00:00
Chris Lattner
87d8429506 Fix miscompilation of MallocBench/espresso which code review pointed out
but apparently didn't make it into the final patch.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31070 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-20 18:20:21 +00:00
Reid Spencer
b83eb6447b For PR950:
This patch implements the first increment for the Signless Types feature.
All changes pertain to removing the ConstantSInt and ConstantUInt classes
in favor of just using ConstantInt.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31063 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-20 07:07:24 +00:00
Devang Patel
6e7dd9db6b While creating mask, use 1ULL instead of 1.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31062 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-20 01:16:56 +00:00
Devang Patel
002e499650 It is OK to remove extra cast if operation is EQ/NE even though source
and destination sign may not match but other conditions are met.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31056 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-19 20:59:13 +00:00
Devang Patel
df308fa7ab Typo Typo.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31055 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-19 19:21:36 +00:00
Devang Patel
85415ed270 Typo.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31054 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-19 19:05:38 +00:00
Devang Patel
6ce890b6ec Fix bug in PR454 resolution. Added new test case.
This fixes llvmAsmParser.cpp miscompile by llvm on PowerPC Darwin.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31053 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-19 18:54:08 +00:00
Reid Spencer
ac5209e5fe Undo Chris' last patch, it caused a regression.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30991 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-16 23:08:08 +00:00
Chris Lattner
02c0e3659f fix a buggy check that accidentally disabled this xform
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30967 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-15 22:42:15 +00:00
Nick Lewycky
078ff41220 Replace custom dispatch code with two uses of InstVisitor. Improves
compile-time performance.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30896 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-12 02:02:44 +00:00
Chris Lattner
c836333c3b Implement SROA of unions with mixed pointers/integers in them. This implements
PR892 and Transforms/ScalarRepl/union-pointer.ll:test2


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30825 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-08 23:53:04 +00:00
Chris Lattner
5b121cc688 Implement Transforms/ScalarRepl/union-pointer.ll:test
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30823 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-08 23:28:04 +00:00
Chris Lattner
867b99f845 add a new SimplifyDemandedVectorElts method, which works similarly to
SimplifyDemandedBits.  The idea is that some operations can be simplified if
not all of the computed elements are needed.  Some targets (like x86) have a
large number of intrinsics that operate on a single element, but pass other
elts through unmodified.  If those other elements are not needed, the
intrinsics can be simplified to scalar operations, and insertelement ops can
be removed.

This turns (f.e.):

ushort %Convert_sse(float %f) {
        %tmp = insertelement <4 x float> undef, float %f, uint 0                ; <<4 x float>> [#uses=1]
        %tmp10 = insertelement <4 x float> %tmp, float 0.000000e+00, uint 1             ; <<4 x float>> [#uses=1]
        %tmp11 = insertelement <4 x float> %tmp10, float 0.000000e+00, uint 2           ; <<4 x float>> [#uses=1]
        %tmp12 = insertelement <4 x float> %tmp11, float 0.000000e+00, uint 3           ; <<4 x float>> [#uses=1]
        %tmp28 = tail call <4 x float> %llvm.x86.sse.sub.ss( <4 x float> %tmp12, <4 x float> < float 1.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > )               ; <<4 x float>> [#uses=1]
        %tmp37 = tail call <4 x float> %llvm.x86.sse.mul.ss( <4 x float> %tmp28, <4 x float> < float 5.000000e-01, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > )               ; <<4 x float>> [#uses=1]
        %tmp48 = tail call <4 x float> %llvm.x86.sse.min.ss( <4 x float> %tmp37, <4 x float> < float 6.553500e+04, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > )               ; <<4 x float>> [#uses=1]
        %tmp59 = tail call <4 x float> %llvm.x86.sse.max.ss( <4 x float> %tmp48, <4 x float> zeroinitializer )          ; <<4 x float>> [#uses=1]
        %tmp = tail call int %llvm.x86.sse.cvttss2si( <4 x float> %tmp59 )              ; <int> [#uses=1]
        %tmp69 = cast int %tmp to ushort                ; <ushort> [#uses=1]
        ret ushort %tmp69
}

into:

ushort %Convert_sse(float %f) {
entry:
        %tmp28 = sub float %f, 1.000000e+00             ; <float> [#uses=1]
        %tmp37 = mul float %tmp28, 5.000000e-01         ; <float> [#uses=1]
        %tmp375 = insertelement <4 x float> undef, float %tmp37, uint 0         ; <<4 x float>> [#uses=1]
        %tmp48 = tail call <4 x float> %llvm.x86.sse.min.ss( <4 x float> %tmp375, <4 x float> < float 6.553500e+04, float undef, float undef, float undef > )           ; <<4 x float>> [#uses=1]
        %tmp59 = tail call <4 x float> %llvm.x86.sse.max.ss( <4 x float> %tmp48, <4 x float> < float 0.000000e+00, float undef, float undef, float undef > )            ; <<4 x float>> [#uses=1]
        %tmp = tail call int %llvm.x86.sse.cvttss2si( <4 x float> %tmp59 )              ; <int> [#uses=1]
        %tmp69 = cast int %tmp to ushort                ; <ushort> [#uses=1]
        ret ushort %tmp69
}

which improves codegen from:

_Convert_sse:
        movss LCPI1_0, %xmm0
        movss 4(%esp), %xmm1
        subss %xmm0, %xmm1
        movss LCPI1_1, %xmm0
        mulss %xmm0, %xmm1
        movss LCPI1_2, %xmm0
        minss %xmm0, %xmm1
        xorps %xmm0, %xmm0
        maxss %xmm0, %xmm1
        cvttss2si %xmm1, %eax
        andl $65535, %eax
        ret

to:

_Convert_sse:
        movss 4(%esp), %xmm0
        subss LCPI1_0, %xmm0
        mulss LCPI1_1, %xmm0
        movss LCPI1_2, %xmm1
        minss %xmm1, %xmm0
        xorps %xmm1, %xmm1
        maxss %xmm1, %xmm0
        cvttss2si %xmm0, %eax
        andl $65535, %eax
        ret


This is just a first step, it can be extended in many ways.  Testcase here:
Transforms/InstCombine/vec_demanded_elts.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30752 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-05 06:55:50 +00:00
Nick Lewycky
f938099aa1 Simplify logic further.
Ensure that we copy KnownProperties before calling visitBasicBlock, else
we may leak properties into blocks where they don't belong.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30705 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 17:36:01 +00:00
Nick Lewycky
a73a654bb4 Simplify, now that predsimplify depends on break-crit-edges.
Fix SwitchInst where dest-block is the same as one of the cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30700 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 15:19:11 +00:00
Nick Lewycky
5c8c5d9ace Move break-crit-edges before the predicate simplifier. Allows us to
optimize in more cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30699 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-03 14:52:23 +00:00
Chris Lattner
e831b9a48b Fix a bug from r1.391 of this file, where we checked the size instead of
the alignment when promoting allocations.  This implements
InstCombine/cast.ll:test32


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30682 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-01 19:40:58 +00:00
Chris Lattner
47811b76ca Eliminate ConstantBool::True and ConstantBool::False. Instead, provide
ConstantBool::getTrue() and ConstantBool::getFalse().


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30665 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-28 23:35:22 +00:00
Chris Lattner
7514437133 set DEBUG_TYPE right
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30623 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-27 04:58:23 +00:00
Nick Lewycky
8f389d8cd4 Style changes only. Remove dead code, fix a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30588 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-23 15:13:08 +00:00
Chris Lattner
9ba46c13bf Fix Transforms/IndVarsSimplify/2006-09-20-LFTR-Crash.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30555 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-21 05:12:20 +00:00
Nick Lewycky
009aa1d485 Don't rewrite ConstantExpr::get.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30552 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-21 01:05:35 +00:00
Nick Lewycky
af2f1fe94e Once we're down to "setcc type constant1, constant2", at least come up
with the right answer.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30550 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-20 23:02:24 +00:00
Nick Lewycky
406fc0cb49 Use a total ordering to compare instructions.
Fixes infinite loop in resolve().


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30540 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-20 17:04:01 +00:00
Andrew Lenharth
4563326472 simplify
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30535 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-20 15:37:57 +00:00
Chris Lattner
b8456460cb We went through all that trouble to compute whether it was safe to transform
this comparison, but never checked it.  Whoops, no wonder we miscompiled
177.mesa!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30511 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-20 04:44:59 +00:00
Evan Cheng
571c15e38a Back out Chris' last set of changes. This breaks 177.mesa and povray somehow.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30505 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-20 01:39:40 +00:00
Evan Cheng
7df4ad9a7f 80 col.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30504 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-20 01:10:02 +00:00
Andrew Lenharth
16d7955eb0 If we have an add, do it in the pointer realm, not the int realm. This is critical in the linux kernel for pointer analysis correctness
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30496 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-19 18:24:51 +00:00
Chris Lattner
733f576d29 implement select.ll:test19-22
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30482 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-19 06:18:21 +00:00
Nick Lewycky
025f4c0fad Walk down the dominator tree instead of the control flow graph. That means
that we can't modify the CFG any more, at least not until it's possible
to update the dominator tree (PR217).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30469 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-18 21:09:35 +00:00
Chris Lattner
6d7ca92bbf Fix an infinite loop building the CFE
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30465 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-18 18:27:05 +00:00
Chris Lattner
e695a3bd32 Implement InstCombine/cast.ll:test31. This speeds up 462.libquantum by 26%.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30456 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-18 05:27:43 +00:00
Chris Lattner
b7363791ed Implement Transforms/InstCombine/shift-sra.ll:test0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30450 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-18 04:31:40 +00:00
Chris Lattner
65b72baa59 Rewrite shift/and/compare sequences to promote better licm of the RHS.
Use isLogicalShift/isArithmeticShift to simplify code.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30448 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-18 04:22:48 +00:00
Chris Lattner
f6bd07c7fe Fix Transforms/InstCombine/2006-09-15-CastToBool.ll and PR913
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30405 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-16 03:14:10 +00:00
Nick Lewycky
668a1d0fd2 Add some more consistency checks.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30305 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-13 19:32:53 +00:00
Nick Lewycky
977be254c6 Fix unionSets so that it can merge correctly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30304 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-13 19:24:01 +00:00
Nick Lewycky
cf2112a0a0 Erase dead instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30298 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-13 18:55:37 +00:00
Chris Lattner
92f7365740 An sinkable instruction may exist with uses, if those uses are in dead blocks.
Handle this.  This fixes PR908 and Transforms/LICM/2006-09-12-DeadUserOfSunkInstr.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30275 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-12 19:17:09 +00:00
Chris Lattner
63d3220dae Fix PR905 and InstCombine/2006-09-11-EmptyStructCrash.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30266 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-11 21:43:16 +00:00
Nick Lewycky
3fc68ccd83 Skip the linear search if the answer is already known.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30251 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-11 17:23:34 +00:00
Chris Lattner
dfa1af0513 Allow tail duplication in more cases, relaxing the previous restriction a
bit.  This fixes Regression/Transforms/TailDup/MergeTest.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30237 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-10 18:17:58 +00:00
Nick Lewycky
dc08cd56c4 Replace EquivalenceClasses with a custom-built data structure. Many common
operations (like findProperties) should be faster, at the expense of
unionSets being slower in cases that are rare in practise.

Don't erase a dead Instruction. This fixes a memory corruption issue.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30235 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-10 02:27:07 +00:00
Chris Lattner
2a86f3b744 Implement Transforms/InstCombine/hoist_instr.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30234 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-09 22:02:56 +00:00
Chris Lattner
8e49e08f4b Turn div X, (Cond ? Y : 0) -> div X, Y
This implements select.ll::test18.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30230 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-09 20:26:32 +00:00
Chris Lattner
e99c623e75 Throttle back tail duplication to avoid creating really ugly sequences of code.
For Transforms/TailDup/if-tail-dup.ll, f.e., it produces:

_foo:
        movl 8(%esp), %eax
        movl 4(%esp), %ecx
        testl $1, %ecx
        je LBB1_2       #cond_next
LBB1_1: #cond_true
        movl $1, (%eax)
LBB1_2: #cond_next
        testl $2, %ecx
        je LBB1_4       #cond_next10
LBB1_3: #cond_true6
        movl $1, 4(%eax)
LBB1_4: #cond_next10
        testl $4, %ecx
        je LBB1_6       #cond_next18
LBB1_5: #cond_true14
        movl $1, 8(%eax)
LBB1_6: #cond_next18
        testl $8, %ecx
        je LBB1_8       #return
LBB1_7: #cond_true22
        movl $1, 12(%eax)
        ret
LBB1_8: #return
        ret

instead of:

_foo:
        movl 4(%esp), %eax
        testl $2, %eax
        sete %cl
        movl 8(%esp), %edx
        testl $1, %eax
        je LBB1_2       #cond_next
LBB1_1: #cond_true
        movl $1, (%edx)
        testb %cl, %cl
        jne LBB1_4      #cond_next10
        jmp LBB1_3      #cond_true6
LBB1_2: #cond_next
        testb %cl, %cl
        jne LBB1_4      #cond_next10
LBB1_3: #cond_true6
        movl $1, 4(%edx)
        testl $4, %eax
        je LBB1_6       #cond_next18
        jmp LBB1_5      #cond_true14
LBB1_4: #cond_next10
        testl $4, %eax
        je LBB1_6       #cond_next18
LBB1_5: #cond_true14
        movl $1, 8(%edx)
        testl $8, %eax
        je LBB1_8       #return
        jmp LBB1_7      #cond_true22
LBB1_6: #cond_next18
        testl $8, %eax
        je LBB1_8       #return
LBB1_7: #cond_true22
        movl $1, 12(%edx)
        ret
LBB1_8: #return
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30158 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-07 21:30:15 +00:00
Nick Lewycky
a3a68bde21 Improve handling of SelectInst.
Reorder operations to remove duplicated work.
Fix to leave floating-point types out of the optimization.
Add tests to predsimplify.ll for SwitchInst and SelectInst handling.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30055 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-02 19:40:38 +00:00
Nick Lewycky
7218c28822 Don't confuse canonicalize and lookup. Fixes predsimplify.reg4.ll. Also
corrects missing optimization opportunity removing cases from a switch.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30009 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-01 03:26:35 +00:00
Nick Lewycky
6e39c7ed1e Properties where both Values weren't in the union (as being equal to
another Value) weren't being found by findProperties.

This fixes predsimplify.ll test6, a missed optimization opportunity.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29991 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-31 00:39:16 +00:00
Nick Lewycky
3947a76e21 Move to using the EquivalenceClass ADT. Removes SynSets.
If a branch's condition has become a ConstantBool, simplify it immediately.
Removing the edge saves work and exposes up more optimization opportunities
in the pass.
Add support for SelectInst.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29970 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-30 02:46:48 +00:00
Devang Patel
4b8f36f106 Do not rely on std::sort and std::erase to get list of unique
exit blocks. The output is dependent on addresses of basic block.

Add and use Loop::getUniqueExitBlocks.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29966 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-29 22:29:16 +00:00
Owen Anderson
d648e140a5 Clean up a bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29950 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-29 06:10:56 +00:00
Nick Lewycky
05450ae128 Add PredicateSimplifier pass. Collapses equal variables into one form
and simplifies expressions. This implements the optimization described
in PR807.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29947 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-28 22:44:55 +00:00
Owen Anderson
59312b19a6 Make LoopUnroll fold excessive BasicBlocks. This results in a significant speedup of
gccas on 252.eon


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29936 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-28 02:09:46 +00:00
Chris Lattner
7f8897f22e eliminate RegisterOpt. It does the same thing as RegisterPass.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29925 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-27 22:42:52 +00:00
Chris Lattner
a4f0b3a084 s|llvm/Support/Visibility.h|llvm/Support/Compiler.h|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29911 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-27 12:54:02 +00:00
Owen Anderson
203572552d Fix a crash related to updating Phi nodes in the original header block. This was
causing a crash in 175.vpr


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29887 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-25 22:13:55 +00:00
Owen Anderson
eb70591492 Add an assertion to check that we're really preserving LCSSA.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29886 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-25 22:12:36 +00:00
Owen Anderson
ac12322710 Reapply the indvars patch, since nothing blew up last night.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29874 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-25 17:41:25 +00:00
Owen Anderson
b397d81f03 Revert my previous patch. Since there are some major changes that went in today,
I'm going to wait to put this in HEAD until tomorrow, so as not to clutter the nightly
tester.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29868 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-25 03:45:57 +00:00
Owen Anderson
8a24749cb8 Specify that indvars actually preserve LCSSA. This has been done for a while, but I
forgot to put in the analysis usage.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29867 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-25 03:32:13 +00:00
Owen Anderson
3b53c4ee91 Implement unrolling of multiblock loops. This significantly improves the
utility of the LoopUnroll pass.

Also, add a testcase for multiblock-loop unrolling.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29859 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-24 21:28:19 +00:00
Reid Spencer
47a53ac726 Fix a grammaro in a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29765 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-18 09:01:07 +00:00
Chris Lattner
df32aac51b Handle single-entry PHI nodes correctly. This fixes PR877 and
Transforms/CondProp/2006-08-14-SingleEntryPhiCrash.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29673 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-14 21:38:05 +00:00
Chris Lattner
7e79b3898d Changes:
1. Update an obsolete comment.
  2. Make the sorting by base an explicit (though still N^2) step, so
     that the code is more clear on what it is doing.
  3. Partition uses so that uses inside the loop are handled before uses
     outside the loop.

Note that none of these changes currently changes the code inserted by LSR,
but they are a stepping stone to getting there.

This code is the result of some crazy pair programming with Nate. :)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29493 91177308-0d34-0410-b5e6-96231b3b80d8
2006-08-03 06:34:50 +00:00
Chris Lattner
ec3c411de8 Minor comment tweaks
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29226 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-20 19:06:16 +00:00
Owen Anderson
10d9d9f27c Add an assertion.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29199 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-19 05:48:45 +00:00
Owen Anderson
2698a08c23 Make LoopUnroll not die on LCSSA Phis. This makes lencod work again.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29198 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-19 05:45:14 +00:00
Owen Anderson
db5b9cfd86 Fix a error that hadn't yet cause any problems, but I'm sure it would have
somewhere down the road.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29197 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-19 03:51:48 +00:00
Evan Cheng
31e773147b Only reuse a previous IV if it would not require a type conversion.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29186 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-18 19:07:58 +00:00
Owen Anderson
c1be492f94 Hopefully the final attempt at making IndVars preserve LCSSA.
This should fix PR 831.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29141 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-14 18:49:15 +00:00
Chris Lattner
262041892d Revert this patch temporarily until PR831 is fixed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29134 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-13 19:05:20 +00:00
Owen Anderson
bcfb913df0 IndVars now (correctly) preserves LCSSA form.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29126 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-12 21:29:14 +00:00
Chris Lattner
ed36b2f1c2 Silence a warning produced in assertions-disabled mode
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29108 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-11 18:31:26 +00:00
Owen Anderson
4a8aadd1f5 Revert my indvars changes because they were breaking things. Unfortunately this
didn't start showing up until after the recent instcombine fixes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29102 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-11 07:25:33 +00:00
Owen Anderson
b64ab8769e Add a comment, and fix a typo that broke the build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29094 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-10 22:15:25 +00:00
Owen Anderson
7e05714a10 Don't indent the entire function.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29093 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-10 22:03:18 +00:00
Chris Lattner
6423d4c64a Recognize 16-bit bswaps by relaxing overconstrained pattern.
This implements Transforms/InstCombine/bswap.ll:test[34].


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29087 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-10 20:25:24 +00:00
Owen Anderson
d1b78a12ef Make instcombine not remove Phi nodes when LCSSA is live.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29083 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-10 19:03:49 +00:00
Chris Lattner
9525528a7d Use hidden visibility to make symbols in an anonymous namespace get
dropped.  This shrinks libllvmgcc.dylib another 67K


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28975 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28 23:17:24 +00:00
Chris Lattner
f4b546110c Shrink libllvmgcc.dylib by another 23K
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28972 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28 22:08:15 +00:00
Owen Anderson
372994be25 Switch to a very conservative heuristic for determining when loop-unswitching
will be profitable.  This is mainly to remove some cases where excessive
unswitching would result in long compile times and/or huge generated code.

Once someone comes up with a better heuristic that avoids these cases, this
should be switched out.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28962 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28 17:47:50 +00:00
Chris Lattner
d228018396 Fix Transforms/InstCombine/2006-06-28-infloop.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28961 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28 17:34:50 +00:00
Chris Lattner
daa2bf9a19 Don't unswitch really large loops even if they are mostly filled with empty
blocks.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28959 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28 16:38:55 +00:00
Andrew Lenharth
b8e604c29a Catch more function pointer casting problems
Remove the Function pointer cast in these calls, converting it to
a cast of argument.
%tmp60 = tail call int cast (int (ulong)* %str to int (int)*)( int 10 )
%tmp60 = tail call int cast (int (ulong)* %str to int (int)*)( uint %tmp51 )


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28953 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-28 01:01:52 +00:00
Owen Anderson
bef8508737 Fix for 2006-06-27-DeadSwitchCase.ll
Be more careful when updating Phi nodes after eliminating dead switch cases.  Fix
proposed by Chris.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28947 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-27 22:26:09 +00:00
Owen Anderson
f9b238e93d De-pessimize the handling of LCSSA Phi nodes in IndVarSimplify. Hopefully this
will make Shootout-C/nestedloop faster.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28924 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-27 02:17:08 +00:00
Chris Lattner
3c80a51cfd random code cleanups, no functionality change
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28914 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-26 19:10:05 +00:00
Owen Anderson
2b67f07d2b Make LoopUnswitch able to unswitch loops with live-out values by taking advantage
of LCSSA.  This results several times the number of unswitchings occurring on
tests such and timberwolfmc, unix-tbl, and ldecod.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28912 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-26 07:44:36 +00:00
Chris Lattner
eb83f4e6cd Fix IndVarsSimplify/2006-06-16-Indvar-LCSSA-Crash.ll, a case where a
"LCSSA" phi node causes indvars to break dominance properties.  This fixes
causes indvars to avoid inserting aggressive code in this case, instead
indvars should be fixed to be more aggressive in the face of lcssa phi's.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28850 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-17 01:02:31 +00:00
Chris Lattner
afe91a5f26 Implement Transforms/InstCombine/bswap.ll, turning common shift/and/or bswap
idioms into bswap intrinsics.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28803 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-15 19:07:26 +00:00
Chris Lattner
f7a1212850 Fix Transforms/LoopUnswitch/2006-06-13-SingleEntryPHI.ll, a loop unswitch
bug exposed by the recent lcssa work.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28779 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-14 04:46:17 +00:00
Owen Anderson
6edf399bb4 Reapply my 6/9 changes. The bug Evan saw no longer occurs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28759 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-12 21:49:21 +00:00
Evan Cheng
b9b2b309d3 Back out Owen's 6/9 changes. They broke MultiSource/Benchmarks/Prolangs-C/bison (and perhaps others).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28747 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-11 09:32:57 +00:00
Owen Anderson
fcaf345c88 Add LCSSA as a requirement for LoopUnswitch, and assert that LoopUnswitch preserves
LCSSA.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28739 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-09 18:40:32 +00:00
Evan Cheng
86c75d3113 RewriteExpr, either the new PHI node of induction variable or the
post-increment value, should be first cast to the appropriated type (to the
type of the common expr). Otherwise, the rewrite of a use based on (common +
iv) may end up with an incorrect type.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28735 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-09 00:12:42 +00:00
Reid Spencer
6296b3cac5 Fix a spello in a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28714 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-07 21:24:10 +00:00
Chris Lattner
18a4af2ba6 Fix a bug in a recent patch. This fixes UnitTests/Vector/Altivec/casts.c on
PPC/altivec


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28698 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-06 22:26:02 +00:00
Chris Lattner
0c1262c9f9 Remove unneeded hook. Patch by Anton K. Thanks!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28664 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-02 19:11:46 +00:00
Chris Lattner
5180b29cb0 Silence a -pedantic warning.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28632 91177308-0d34-0410-b5e6-96231b3b80d8
2006-06-01 17:16:21 +00:00
Chris Lattner
13d4ab4009 Swap the order of operands created here. For +&|^, the order doesn't matter,
but for sub, it really does!  Fix fixes a miscompilation of fibheap_cut in
llvmgcc4.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28600 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-31 21:14:00 +00:00
Chris Lattner
b4db97ffa7 Implement Transforms/InstCombine/store.ll:test2.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28503 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-26 19:19:20 +00:00
Chris Lattner
7b2e2792d6 Transform things like (splat(splat)) -> splat
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28490 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-26 00:29:06 +00:00
Chris Lattner
863bcff183 Introduce a helper function that simplifies interpretation of shuffle masks.
No functionality change.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28489 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-25 23:48:38 +00:00
Chris Lattner
01575b7315 Turn (cast (shuffle (cast)) -> shuffle (cast) if it reduces the # casts in
the program.  This exposes more opportunities for the instcombiner, and implements
vec_shuffle.ll:test6


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28487 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-25 23:24:33 +00:00
Chris Lattner
73fa49d697 extract element from a shuffle vector can be trivially turned into an
extractelement from the SV's source.  This implement vec_shuffle.ll:test[45]


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28485 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-25 22:53:38 +00:00
Chris Lattner
01859e8853 Silence a bogus gcc warning
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28422 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-20 23:14:03 +00:00
Evan Cheng
3e246dd084 Backing out last check-in for now. It's causing an infinite loop gccas lencode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28284 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-14 06:46:03 +00:00
Chris Lattner
033aaaf451 Add/Sub/Mul are safe to promote here as well. Incrementing a single-bit
bitfield now gives this code:

_plus:
        lwz r2, 0(r3)
        rlwimi r2, r2, 0, 1, 31
        xoris r2, r2, 32768
        stw r2, 0(r3)
        blr

instead of this:

_plus:
        lwz r2, 0(r3)
        srwi r4, r2, 31
        slwi r4, r4, 31
        addis r4, r4, -32768
        rlwimi r2, r4, 0, 0, 0
        stw r2, 0(r3)
        blr

this can obviously still be improved.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28275 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-13 02:16:08 +00:00
Chris Lattner
70074e00a2 Implement simple promotion for cast elimination in instcombine. This is
currently very limited, but can be extended in the future.  For example,
we now compile:

uint %test30(uint %c1) {
        %c2 = cast uint %c1 to ubyte
        %c3 = xor ubyte %c2, 1
        %c4 = cast ubyte %c3 to uint
        ret uint %c4
}

to:

_xor:
        movzbl 4(%esp), %eax
        xorl $1, %eax
        ret

instead of:

_xor:
        movb $1, %al
        xorb 4(%esp), %al
        movzbl %al, %eax
        ret

More impressively, we now compile:

struct B { unsigned bit : 1; };
void xor(struct B *b) { b->bit = b->bit ^ 1; }

To (X86/PPC):

_xor:
        movl 4(%esp), %eax
        xorl $-2147483648, (%eax)
        ret
_xor:
        lwz r2, 0(r3)
        xoris r2, r2, 32768
        stw r2, 0(r3)
        blr

instead of (X86/PPC):

_xor:
        movl 4(%esp), %eax
        movl (%eax), %ecx
        movl %ecx, %edx
        shrl $31, %edx
        # TRUNCATE movb %dl, %dl
        xorb $1, %dl
        movzbl %dl, %edx
        andl $2147483647, %ecx
        shll $31, %edx
        orl %ecx, %edx
        movl %edx, (%eax)
        ret

_xor:
        lwz r2, 0(r3)
        srwi r4, r2, 31
        xori r4, r4, 1
        rlwimi r2, r4, 31, 0, 0
        stw r2, 0(r3)
        blr

This implements InstCombine/cast.ll:test30.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28273 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-13 02:06:03 +00:00
Chris Lattner
8c8c66a8cd Refactor some code, making it simpler.
When doing the initial pass of constant folding, if we get a constantexpr,
simplify the constant expr like we would do if the constant is folded in the
normal loop.

This fixes the missed-optimization regression in
Transforms/InstCombine/getelementptr.ll last night.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28224 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-11 17:11:52 +00:00
Chris Lattner
f4f5a77fdc Two changes:
1. Implement InstCombine/deadcode.ll by not adding instructions in unreachable
   blocks (due to constants in conditional branches/switches) to the worklist.
   This causes them to be deleted before instcombine starts up, leading to
   better optimization.

2. In the prepass over instructions, do trivial constprop/dce as we go.  This
   has the effect of improving the effectiveness of #1.  In addition, it
   *significantly* speeds up instcombine on test cases with large amounts of
   constant folding code (for example, that produced by code specialization
   or partial evaluation).  In one example, it speeds up instcombine from
   0.0589s to 0.0224s with a release build (a 2.6x speedup).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28215 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-10 19:00:36 +00:00
Chris Lattner
33a6113995 Move some code around.
Make the "fold (and (cast A), (cast B)) -> (cast (and A, B))" transformation
only apply when both casts really will cause code to be generated.  If one or
both doesn't, then this xform doesn't remove a cast.

This fixes Transforms/InstCombine/2006-05-06-Infloop.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28141 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-06 09:00:16 +00:00
Chris Lattner
581a7ad7c9 Fix an infinite loop compiling oggenc last night.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28128 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-05 20:51:30 +00:00
Chris Lattner
6fc205fc44 Implement InstCombine/cast.ll:test29
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28126 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-05 06:39:07 +00:00
Chris Lattner
e3158308e0 Fix Transforms/InstCombine/2006-05-04-DemandedBitCrash.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28101 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-04 17:33:35 +00:00
Chris Lattner
94046b4d10 Fix InstCombine/2006-04-28-ShiftShiftLongLong.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28019 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-28 22:21:41 +00:00
Chris Lattner
ae74f55552 Fix Transforms/Reassociate/2006-04-27-ReassociateVector.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28007 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-28 04:14:49 +00:00
Chris Lattner
d929f06f4d Add support for inserting undef into a vector. This implements
Transforms/InstCombine/vec_insert_to_shuffle.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27997 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-27 21:14:21 +00:00
Chris Lattner
df4f226cdc Fix Transforms/ScalarRepl/2006-04-20-PromoteCrash.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27912 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-20 20:48:50 +00:00
Andrew Lenharth
7a31b975fb Make code match cvs commit message :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27881 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-20 15:41:37 +00:00
Andrew Lenharth
8117f9a40e If we can convert the return pointer type into an integer that IntPtrType
can be converted to losslessly, we can continue the conversion to a direct call.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27880 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-20 14:56:47 +00:00
Chris Lattner
fd6bdf0b0f Turn x86 unaligned load/store intrinsics into aligned load/store instructions
if the pointer is known aligned.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27781 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-17 22:26:56 +00:00
Chris Lattner
7f6cc0ccb5 Fix a bug in the 'shuffle(undef,x,mask) -> shuffle(x, undef,mask')' xform
Make the insert/extract elt -> shuffle code more aggressive.

This fixes CodeGen/PowerPC/vec_shuffle.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27728 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-16 00:51:47 +00:00
Chris Lattner
706126dac1 Canonicalize shuffle(undef,x,mask) -> shuffle(x, undef,mask').
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27727 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-16 00:03:56 +00:00
Chris Lattner
efb47359ff significant cleanups to code that uses insert/extractelt heavily. This builds
maximal shuffles out of them where possible.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27717 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-15 01:39:45 +00:00
Chris Lattner
de6df88529 Teach scalarrepl to promote unions of vectors and floats, producing
insert/extractelement operations.  This implements
Transforms/ScalarRepl/vector_promote.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27710 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-14 21:42:41 +00:00
Reid Spencer
ad2072643a Get rid of a signed/unsigned compare warning.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27625 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-12 19:28:15 +00:00
Chris Lattner
9fb9213083 Turn casts into getelementptr's when possible. This enables SROA to be more
aggressive in some cases where LLVMGCC 4 is inserting casts for no reason.

This implements InstCombine/cast.ll:test27/28.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27620 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-12 18:09:35 +00:00
Chris Lattner
389a6f58f7 Implement vec_shuffle.ll:test3
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27573 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-10 23:06:36 +00:00
Chris Lattner
a844fc4cb0 Implement InstCombine/vec_shuffle.ll:test[12]
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27571 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-10 22:45:52 +00:00
Chris Lattner
543abdf1d0 Add supprot for shufflevector
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27513 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-08 01:19:12 +00:00
Chris Lattner
e2ed057562 Lower vperm(x,y, mask) -> shuffle(x,y,mask) if mask is constant. This allows
us to compile oh-so-realistic stuff like this:

 vec_vperm(A, B, (vector unsigned char){14});

to:
        vspltb v0, v0, 14

instead of:

        vspltisb v0, 14
        vperm v0, v2, v1, v0


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27452 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-06 19:19:17 +00:00
Chris Lattner
4132afb0d2 vector casts of casts are eliminable. Transform this:
%tmp = cast <4 x uint> %tmp to <4 x int>                ; <<4 x int>> [#uses=1]
        %tmp = cast <4 x int> %tmp to <4 x float>               ; <<4 x float>> [#uses=1]

into:

        %tmp = cast <4 x uint> %tmp to <4 x float>              ; <<4 x float>> [#uses=1]


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27355 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 05:43:13 +00:00
Chris Lattner
a1c3538537 Allow transforming this:
%tmp = cast <4 x uint>* %testData to <4 x int>*         ; <<4 x int>*> [#uses=1]
        %tmp = load <4 x int>* %tmp             ; <<4 x int>> [#uses=1]

to this:

        %tmp = load <4 x uint>* %testData               ; <<4 x uint>> [#uses=1]
        %tmp = cast <4 x uint> %tmp to <4 x int>                ; <<4 x int>> [#uses=1]


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27353 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 05:37:12 +00:00
Chris Lattner
82ed58f9c4 Turn altivec lvx/stvx intrinsics into loads and stores. This allows the
elimination of one load from this:

int AreSecondAndThirdElementsBothNegative( vector float *in ) {
#define QNaN 0x7FC00000
const vector unsigned int testData = (vector unsigned int)( QNaN, 0, 0, QNaN );
vector float test = vec_ld( 0, (float*) &testData );
return ! vec_any_ge( test, *in );
}

Now generating:

_AreSecondAndThirdElementsBothNegative:
        mfspr r2, 256
        oris r4, r2, 49152
        mtspr 256, r4
        li r4, lo16(LCPI1_0)
        lis r5, ha16(LCPI1_0)
        addi r6, r1, -16
        lvx v0, r5, r4
        stvx v0, 0, r6
        lvx v1, 0, r3
        vcmpgefp. v0, v0, v1
        mfcr r3, 2
        rlwinm r3, r3, 27, 31, 31
        xori r3, r3, 1
        cntlzw r3, r3
        srwi r3, r3, 5
        mtspr 256, r2
        blr


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27352 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-02 05:30:25 +00:00
Chris Lattner
ae1ab3989c Fix InstCombine/2006-04-01-InfLoop.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27330 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-01 22:05:01 +00:00
Chris Lattner
64daab56c2 Fold A^(B&A) -> (B&A)^A
Fold (B&A)^A == ~B & A

This implements InstCombine/xor.ll:test2[56]


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27328 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-01 08:03:55 +00:00
Chris Lattner
6e6b0da303 If we can look through vector operations to find the scalar version of an
extract_element'd value, do so.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27323 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-31 23:01:56 +00:00